Given a Strategic Asset Allocation (the proportions/weights of the various asset classes), then you must find the appropriate vehicle to implement each asset class in the portfolio. For each asset class one can choose a passive (composition same as the index of the asset class) or an active approach (whereby composition of each asset class is different from that corresponding of the index).
As discussed in the asset allocation section, in the past people tended to stay close to home, especially Americans who were almost completely invested in U.S. based securities. In Canada, even if investors were somewhat better in allocating a larger proportion of their assets to foreign securities, the problem could be more severe. Canadian securities represent 2-3% of the global investable assets, whereas U.S. securities represent about 50% of the global assets. So an American investing 100% in the U.S may still be more diversified than a Canadian investing only 80% in Canada. In addition for the overall portfolio, even while maintaining the SAA prescribed asset type/class allocations, you could still choose weights different from the geographical weights in the world index, while still meeting the asset class allocations. As indicated earlier, asset allocation has the single greatest explanatory power in the variability between portfolio returns. There is little supporting historical evidence that fund managers can beat the various indexes in a sustained manner after fees, transaction costs and taxes. Not only are the odds for an individual manager against exceeding the index return (which is the average of all investors before fees, transaction costs and taxes), but you the investor would have to have chosen one of the few winners! This has not proven to be a successful approach used by investors. What’s even worse, is that studies have shown that investors get even lower return than the mutual funds they invest in, as they then to look for funds which were winners in the past and then sell after they underperformed for a couple of years.
A better strategy may be to stay with passive vehicles in each asset class and perhaps customize the SAA (and even the TAA). A passive implementation via a mutual fund (for U.S. based investors) or an ETF for Canadian investors may be a workable choice. U.S. based index mutual funds will have very low management fees of 0.1% for domestic to about 0.3% for emerging market funds at the low cost family of Vanguard.
The simplest way to implement this would be to invest via a low cost ‘Life-cycle Fund’ mutual fund (many are now available in the U.S.). For example if you are planning to retire in 2030 you could choose Vanguard’s Target Retirement 2030 Fund. This fund in 2006 allocates about 70% and 13% to the U.S. total stock and bond market indexes and about 10%, 5% and 3% to European, Pacific and Emerging market stock indexes. Over time the asset mix evolves gradually toward what Vanguard considers a suitable asset mix as you approach retirement. For example the in 2006 the Target Retirement 2005 fund is composed of about 41%, 35% and 13% Total Bond Market, Total Stock Market indexes and Inflation Protected Securities funds respectively; additionally about 5%, 3%, 1% and 1% are allocated to European, Pacific, Emerging Market and Money Market funds. Note, how the risk of the 2005 portfolio is significantly lower than the 2030 portfolio, by virtue of much higher bond content and lower foreign stock content. The management fees associated with these funds is the blended cost of the various index funds used; the 2030 one being 0.21%. I am not aware of any such funds in Canada, and there were any, judging by other Canadian funds, they would have much higher management fees. Nevertheless the basic approach can be implemented in Canada in almost as cost effective way using ETFs.
A passive approach is generally intended to replicate some index. The cost of implementing a passive investment strategy is lower than active one because of the lower management effort/fees, lower turnover and thus lower transaction costs and taxes. The passive approach can be accessed for many indexes via mutual funds or ETFs. The ETFs which are listed on public stock exchanges have the additional advantage of being tradable throughout the day, while the exchanges are open.
For a Canadian investor ETFs tend to be the cheaper vehicles for implementing a well diversified portfolio, than using mutual funds. For more obscure emerging market applications, when ETFs were not available I may have chosen a CEF, Closed End Fund. CEFs which are just like mutual funds, except the manager does not accept additional money into the fund; investors must purchase units on the open market.
It would not be unfair to say that U.S. mutual funds, Canadian mutual funds are on the average 1% and 2% more expensive that corresponding categories of ETFs; in fact that would be generous when considering the recent academic study (Khorana, Servaes, Tufano) naming Canadian funds as the world’s most expensive ones. So if a stock index returns annually 9.5% and assuming an ETF fee drag of 0.5%, then implementing that asset class via ETFs, U.S. mutual fund and Canadian mutual fund would return annually 9%, 8% and 7% respectively. One may argue that a manager may be able to earn returns to overcome the impact of fee differentials; however the historical data does not show that there are many managers who can do this on a sustained basis or that if there were such managers, we could predict who they were before the fact.
So after 30 years $1,000 would become $13,268, $10,063 and $7,612, at 9%, 8% and 7% respectively (i.e. after 30 years there asset loss is about 24% for each 1% reduction in return due to fees). Also, after 30 years of $1,000/year investments, the assets would grow to approximately $136,000, $113,000 and $94,000, at 9%, 8% and 7% respectively (i.e. after 30 years there asset loss is about 17% for each 1% reduction in return due to fees) . An investor saving for retirement would find it inadvisable to ignore the impact of such a cost difference.
In the following, I have created a matrix in the asset type and geographical dimensions representing ETFs that a Canadian investor may consider for implementing such a diversified portfolio:
These are ETFs and CEFs (Closed End Funds) that I am currently using or have used in the past in my portfolio. A couple of point worth mentioning are that additional information on these ETFs and CEFs are available at eftconnect.com (GIM,MSD,GLD) or at sponsoring institutions like ishares.com (EFA,EEM), vanguard.com (VTI), and ishares.ca (XIU,XBB). Management Expense Ratios for stock vehicles range from 0.07 to 0.35 to 0.75% for VTI, EFA and EEM respectively; in case of bond vehiclesMER is 0.30 for Canadian ETF XBB going to 0.8 and 1.36% for GIM and MSD which are CEFs. There usually multiple implementation vehicles available for each asset class. When comparing among implementation vehicles, in addition toMER, you must consider for example: liquidity, differences in coverage/diversification of the specific index, currency hedging, discount/premium for CEFs.
My preference has been to use ETFs which were not currency hedged back into Canadian dollars because I felt that since I had no clue which way CAD was going and I was spending in both in and out of Canada, staying unhedged actually enhanced my diversification, and I didn’t pay hedging costs. Should you wish to have the international funds hedged you could get at times the equivalent Canadian ishares funds like XIN instead of EFA; XIN hedges EFA back into CAD.
I did not use in this example for a Canadian investor any commodity funds, since Canadians already have relatively high commodity exposure by just buying the Canadian market. IGE (representing Goldman Sachs Natural Resources Index) could be a candidate for the commodity asset class, but it has a very high proportion of oil and oil services content (almost 80%). GLD is a good replication of the intended gold in the portfolio.
There are couple of closed end funds, TDF and IFN, that some use as a subset of the emerging market asset class to cover the China and India stories, respectively.
And finally TIP to replicate performance of inflation protected U.S. Treasury (TIPS), which experts feel are an asset class in themselves and should be included in the asset allocation.
A benchmark is the performance of some reference portfolio. For example for the above investor we could choose a benchmark composed of 1/3:1/3:1/3 Canadian: U.S.: EAFE stock indexes and 2/3:1/3 Canadian: International bond indexes, with stocks to bonds weight of 60:40 overall.
Portfolio Monitoring and Rebalancing
Ideally you’d want to monitor your portfolio for overall performance and risk. The simplest way to insure that as market prices change and you don’t deviate significantly from your target asset allocation, and at the same time to increase the likelihood that you replace expensive assets with cheaper ones is to rebalance the portfolio if it strays outside some predetermined range. Some argue that this should be done on a regular schedule (quarterly, semiannually or annually) while others advocate that this be done anytime when asset allocation drift outside the range. There are all kinds of variations to this intended to minimize risk, maximize performance and/or minimize transaction costs